Implantable neuromodulation systems have proven therapeutic in a wide variety of diseases and disorders. Pacemakers and Implantable Cardiac Defibrillators (ICDs) have proven highly effective in the treatment of a number of cardiac conditions (e.g., arrhythmias). Spinal Cord Stimulation (SCS) systems have long been accepted as a therapeutic modality for the treatment of chronic pain syndromes, and the application of tissue stimulation has begun to expand to additional applications such as angina pectoralis and incontinence. Deep Brain Stimulation (DBS) has also been applied therapeutically for well over a decade for the treatment of refractory chronic pain syndromes, and DBS has also recently been applied in additional areas such as movement disorders and epilepsy. Further, in recent investigations, Peripheral Nerve Stimulation (PNS) systems have demonstrated efficacy in the treatment of chronic pain syndromes and incontinence, and a number of additional applications are currently under investigation. Furthermore, Functional Electrical Stimulation (FES) systems have been applied to restore some functionality to paralyzed extremities in spinal cord injury patients.
Each of these implantable neuromodulation systems typically includes at least one neuromodulation lead implanted at the desired modulation site and an neuromodulation device, such as an implantable pulse generator (IPG), implanted remotely from the modulation site, but coupled either directly to the neuromodulation lead(s), or indirectly to the neuromodulation lead(s) via one or more lead extensions. Thus, electrical pulses can be delivered from the neuromodulation device to the electrodes carried by the neuromodulation lead(s) to modulate a volume of tissue in accordance with a set of modulation parameters and provide the desired efficacious therapy to the patient. The neuromodulation system may further comprise a handheld remote control (RC) to remotely instruct the neuromodulator to generate electrical modulation pulses in accordance with selected modulation parameters. The RC may, itself, be programmed by a technician attending the patient, for example, by using a Clinician's Programmer (CP), which typically includes a general purpose computer, such as a laptop, with a programming software package installed thereon.
Electrical modulation energy may be delivered from the neuromodulation device to the electrodes in the form of an electrical pulsed waveform. Thus, electrical modulation energy may be controllably delivered to the electrodes to modulate neural tissue. The configuration of electrodes used to deliver electrical pulses to the targeted tissue constitutes an electrode configuration, with the electrodes capable of being selectively programmed to act as anodes (positive), cathodes (negative), or left off (zero). In other words, an electrode configuration represents the polarity being positive, negative, or zero. Other parameters that may be controlled or varied include the amplitude, width, and rate of the electrical pulses provided through the electrode array. Each electrode configuration, along with the electrical pulse parameters, can be referred to as a “modulation parameter set.”
With some neuromodulation systems, and in particular, those with independently controlled current or voltage sources, the distribution of the therapeutic electrical current between the electrodes (including the case of the neuromodulation device, which may act as an electrode) may be varied such that the current is supplied via numerous different electrode configurations. In different configurations, the electrodes may provide current or voltage in different relative percentages of positive and negative current or voltage to create different electrical current distributions (i.e., fractionalized electrode configurations).
More pertinent to the present inventions, a neuromodulation device may include one or more current sources/sinks that are configured to supply/receive therapeutic electrical current to/from the electrodes. For example, as shown in FIG. 1, a basic output current source 1 and a corresponding output current sink 2 used to deliver electrical energy to tissue exemplified generically as a load resistance R will be described. The output current source 1 includes a current generator 3, digital-to-analog circuitry (DAC) 4, and a selection transistor 5. Likewise, the output current sink 2 includes a current generator 6, a DAC 7, and a selection transistor 8.
Each of the current generators 3, 6 includes transistors M1, M3 each configured for generating a reference current Irref. Each of the DACs 4, 7 is configured for scaling the reference current Iref using a parallel number N of transistors M2, M4. It should be appreciated that each of the transistors M1/M3 and transistors M2/M4 can be considered current mirrors. The transistors M1, M3 in the output current source 1 are P-type transistors, and thus, the DAC 4 can be considered a PDAC, and similarly, the output current source 1 can be considered PDAC circuitry. In contrast, the transistors M2, M4 in the output current sink 2 are N-type transistors, and thus, the DAC 7 can be considered an NDAC, and similarly, the output current sink 2 can be considered NDAC circuitry. Without a full discussion of transistor physics, one skilled in the art will recognize that use of transistors of such polarities is sensible, given that the output current source 1 will be tied to a positive voltage (V+, referred to herein as the “compliance voltage”), while the output current sink 2 will be tied to a more negative voltage, such as ground. A “ground voltage” as used herein should be understood as any reference voltage with respect to the compliance voltage.
Each of the selection transistors 5, 8 selects the number of output stages M2, M4 to be activated in the respective DACs 4, 7 in response to the input of a digital signal. Therefore, the DAC 4 may scale the reference current Iref by the selected number j to source an output current Iout equal to j*Iref to electrode Ex, and the DAC 7 may scale a selection transistor 5 by the selected number k to sink an input current in equal to k*Iref from electrode Ey. Thus, the output current source 1 and output current sink 2 are generally digitally controllable by the selection transistors 5, 8 to respectively generate the output current Iout and input current Iin. If the electrodes Ex, Ey are the only electrodes utilized by the neurostimulator, the current Iout at Ex will be equal to the current Iin at Ey. However, as is typical, more than two electrodes may be used, in which case the output current sourced to a particular electrode may not be equal to the output current sunk into another electrode. In any case, the sum of the output current Iout sourced by any number of electrodes will be equal to the sum of the input current Iin sunk to any number of electrodes
As just alluded to, a neuromodulator typically operates with several electrodes, and the various current sources and sinks can be controlled to source or sink current to any particular electrode as is efficacious for treating a particular patient. Different output source/sink architectures can be used in a neuromodulation device. For example, each electrode can be coupled to dedicated PDAC/NDAC circuitry, which allows the electrode to either operate as a current source or a current sink, as described in U.S. Pat. No. 6,181,996, which is expressly incorporated herein by reference. As another example, PDAC/NDAC circuitry can be selectively coupled to any of the electrodes via a low-impedance switching matrix, as described in U.S. Pat. No. 6,516,227, which is expressly incorporated herein by reference. As still another example, instead of using discrete PDAC and NDAC blocks that services the various electrodes, the PDAC and NDAC circuitry is effectively distributed such that any of a number of current mirrors can be coupled to any of the electrodes, as described in U.S. patent application Ser. No. 11/177,503, which is expressly incorporated herein by reference.
Regardless of the current source/sink architectures used, all generally have similar current output path characteristics. That is, referring back to FIG. 1, the current output paths in each architecture comprises, at a minimum, a current source output transistor (or transistors if paralleled for current gain) 3, a selection transistor 5 to control the flow of the current source transistor(s) 3, the load resistance R, a current sink transistor (or transistors if paralleled for current gain) 6, and a selection transistor 7 to control the flow of the current sink transistor(s) 6. Each of these elements has some resistance, and hence some amount of the compliance voltage V+ will be dropped across these elements when current is flowing through the load resistance R. Specifically, the compliance voltage V+ will equal VDS1+VR+VDS2, where VDS1 is the drain-to-source voltage drop across the current source transistor(s) 3 and the selection transistor 4, and VDS2 is the drain-to-source voltage drop across the current sink transistor(s) 6 and the selection transistor 7, and VR equals the voltage drop across the load resistance R.
It should be appreciated that the M1/M3 and M2/M4 current mirrors require that transistors M1 and M2 operate in saturation mode, such that the channels of the transistors are in “pinch off,” as illustrated in FIG. 2. When in the saturation mode, the output current Iout is proportional to the gate voltage of the transistors M1 or M2, but does not depend upon the drain voltage to the first order. However, to keep the transistors M1 and M2 in the saturation mode, a certain drain-to-source voltage VDS has to be satisfied for each transistor.
What this means in the context of the output current circuitry of FIG. 1 is that the circuit can operate properly over a range of compliance voltages V+. For example, suppose a suitable therapy for a patient suggests that a current of Iout=5 mA should be passed between electrodes Ex and Ey. Suppose further that the load resistance R equals 800 ohms. When the current of 5 mA is passed through the load resistance R, a voltage VR=4V will build up across the resistance load R (V=I*R). Suppose further for simplicity that the minimum drain-to-source voltage to keep the output transistors M1 and M2 in saturation equals 1V when the effects of the selection transistors 4, 7 are included. The actual value can be different, but is chosen as 1V for ease of illustration. To provide this current, a minimum compliance voltage V+ of at least 6V would be needed; if V+<6V, the circuitry will be unable to produce the desired amount of current.
The compliance voltage V+ could be higher than 6V while still producing the proper amount of current. For example, suppose for the same example that the compliance voltage V+ is 8V. In this case, the circuitry is still capable of providing the 5 mA current, and the load (which doesn't change) will still drop 4V at that current. What this means is that the remainder of the compliance voltage must be dropped across the output transistors M1 and M2 as well as their associated selection transistors 4, 7, e.g., 2V if the source and sink are matched.
However, running the circuit in this example with an 8V compliance voltage is not efficient. While circuit performance is the same at both 6V and 8V, i.e., both are capable of generating a 5 mA current. At 6V, only 30 mW of power (P=I*V) will be drawn, while at 8V, 40 mW of power will be drawn. In other words, 10 mW of power is needlessly dropped across the output transistors M1, M2 and their selection transistors 4, 7. This waste of power is regrettable in the context of an implantable medical device, such as an IPG, which requires a source of energy either supplied by a battery or an external charging source. Therefore, it is important to minimize circuit operation that would otherwise needlessly drain the battery and cause the IPG to cease functioning, or needlessly require the patient to more frequency recharge the battery.
Unfortunately, it is difficult to design the compliance voltage to an optimum level. Depending on the electrodes that are activated, the magnitude of the current required for efficient therapy for a given patient, and the resistance of the patient's flesh, an optimal compliance voltage from the vantage point of power conservation is variable. As such, mechanisms have been designed into prior art neuromodulation systems that adjust the compliance voltage each time the programmed electrical current amplitude or electrode combination is changed by the user. Although the compliance voltage can theoretically be adjusted at a rapid rate (e.g., every minute) to compensate for potential changes in the tissue environment, thereby ensuring that the current source/sink circuitry continues to function properly to provide the current at the programmed amplitude in response to these tissue impedance changes, the compliance voltage adjustment requires bursts of high power drain, and may consume significant amounts of energy. Thus, performing too many compliance voltage adjustments will waste energy. In extreme cases, constant compliance voltage adjustments not only creates high system power consumption, but also prevents the IPG from performing other tasks. As such, a fixed compliance voltage margin (e.g., 12%) is built into the adjusted compliance voltage to ensure that the delivered therapy is not compromised without having to continually make compliance voltage adjustments.
This compliance voltage margin, of course, represents wasted energy, and if the tissue environment has stabilized over a period of time, the compliance margin may be unnecessarily too large. Furthermore, in the context of some therapeutic applications, such as SCS, the change in the tissue impedance is rather slow relative to the frequency at which the amplitude and/or electrode combination, and thus the compliance voltage, is adjusted. As such, a moderate compliance voltage margin, such 12%, will be sufficient to compensate for the tissue impedance changes between compliance voltage adjustments.
However, in other therapeutic applications, such as DBS, it has been discovered that the impedance of the tissue (in the case of DBS, brain tissue) varies greatly over both the long term and the short term. In particular, there have been a number of DBS impedance data sets from animal trials and limited human experiments suggesting that brain tissue impedance tends to vary significantly during both the long term and short term.
For example, it has been demonstrated that the tissue impedance of brain tissue measured from a neuromodulation lead rapidly increases during the first four weeks of implantation (in this case, about 40%), gradually decreases during the next eight weeks after implantation (in this case, about −40%), and then stabilizes thereafter, as shown in FIG. 3. If the compliance voltage is left unchanged after implantation, the therapy will be significantly compromised (under compliance) two weeks after implantation until the impedance subsequently drops to a level where the compliance voltage is sufficient. Even if the amplitude and/or electrode combination, and thus the compliance voltage, is adjusted at least one time during this period, thereby at least partially compensating for the change in impedance over the long term, the compliance voltage margin, which translates to a higher compliance voltage, will be relatively large when the tissue impedance stabilizes, thereby unnecessarily wasting energy.
It has also been demonstrated that the tissue impedance of brain tissue measured from a neuromodulation lead rapidly increases from a baseline level to a peak during the first ten minutes of electrical energy delivery (in this case, about 30%), rapidly decreases during the next ten minutes of electrical energy delivery (in this case, about −30%), gradually decreases during the next forty minutes (in this case, about −15%), and then stabilizes thereafter, as shown in FIG. 4. Because it is unlikely that the amplitude value or electrode combination would be adjusted during the initial sixty minute period of therapy, or at least at the rate at which compliance voltage adjustments can effectively compensate for the impedance changes, the therapy will be significantly compromised during the first twenty minutes (during the rapid increase of the tissue impedance to the peak, and the rapid decrease of the tissue impedance to the baseline level) and will considerably waste energy for the remainder of the therapy session (during the gradual decrease of the tissue impedance from the baseline level).
It can be appreciated from the foregoing that an improved technique for effectively and efficiently adjusting the compliance voltage of a neuromodulation device designed to deliver a constant current is needed.